Why digitalization in mechanical engineering often remains invisible

The well-known “productivity paradox” describes the finding that in mechanical engineering, productivity remained weak despite revenue growth after the financial crisis. After the starting signal for Industry 4.0 at Hannover Messe in 2012, many companies hesitated to implement it. Transformation costs money, and the benefits often only become apparent with a time lag. Then came Covid, supply chain crises, and economic downturns. Stagnating productivity and rising unit labor costs reinforced the impression of a structural lag. This has entrenched the thesis that German industry has slept through digitalization.

Hiding your light under a bushel – why digitalization in mechanical engineering often remains invisible

According to an analysis by the Fraunhofer Institute for Systems and Innovation Research (ISI), only around 30% of industrial companies in Germany systematically use Industry 4.0 technologies in production. However, this figure captures the level of adoption, but not the maturity, quality, and economic impact of the leading applications and of the companies that set the benchmarks in industry and thus drive development. In mechanical engineering in particular, digitalization is evident primarily in processes and products. It is therefore less visible than in platform-driven business models, says Oliver Schöllenhammer, head of the digitalization and AI business unit at the Fraunhofer Institute for Manufacturing Engineering and Automation (IPA). Many companies are effectively hiding their progress under a bushel. “More is being done than is visible,” he says.

An example of underestimating oneself: Only 3% of respondents in an IPA survey agreed with the statement that their production system provides the basis for radical innovations. In reality, however, many companies were already using networked machines, data-based applications, and automated processes. This suggests that some of the digital progress is not even classified internally as strategic transformation. Of course, there are laggards, but they exist everywhere.

The productivity paradox can also be explained by the fact that many technologies had long been available but were not yet mature enough in practice to generate broad productivity effects. Only today’s convergence of AI, automation, connectivity, and computing power is turning them into solutions with real impact. German companies need to learn from the rapid pace of development, especially in China; they can adapt this speed. And they should avoid a typical mistake: thinking from the perspective of technologies - more computing power, machine learning, virtualization - instead of starting from the target system the company actually wants to achieve. 

Value creation instead of tools - why digitalization in mechanical engineering is more than technology

“In mechanical and plant engineering, digitalization is still too often reduced to sensors, the cloud, or individual software solutions,” says Prof. Claus Oetter, managing director of the Software and Digitalization Association at VDMA. In fact, it is about productivity, new value creation, resilience, and technological sovereignty. Digitalization is successful when companies continue to develop machines, processes, and business models in a software-centric way and turn data into tangible benefits for customers.

This fits with the fact that mechanical engineering is gradually moving away from pure product business. Networked machines create the basis for remote monitoring, condition monitoring, virtual services, or optimized maintenance. Building on this, data-based business models such as subscriptions, pay-per-use, or equipment-as-a-service are emerging. Machines are not only sold, but billed based on usage; at the same time, they supply status and process data that manufacturers can use to improve availability, service quality, and margins. Oetter therefore describes digital business models as the key to new value creation opportunities. The prerequisites are networked machines, clear customer value, a robust pricing logic, strategic integration, and change management.

HAWE Hydraulik is already using production-related digitalization in ongoing operations. Machines of different generations are connected, and production, quality, and status data are systematically recorded and used for planning, control, and analysis. The next step is the operational use of live data. When a machine detects a new tool, an inspection process is triggered automatically and a completed inspection plan is immediately made available to the operator. Manual steps are eliminated, and the process becomes faster and more robust. “Digitalization is not an off-the-shelf product. It only works if it is consistently aligned with the company’s real processes and needs,” says Dr.-Ing. Jens Folmer, manager digital transformation at HAWE. Johannes Simon, team lead operational excellence, adds: “Many companies now have a lot of data – the real challenge is to make this data quickly and meaningfully usable in day-to-day production.” For HAWE, the main challenge lies in linking OT and IT. That is why, in addition to classic IT, the company is increasingly relying on low code, citizen development, and empowering the specialist departments. Trumpf also demonstrates how digitalization is created directly in the product: with the AI-supported “Cutting Assistant,” cutting parameters for laser cutting systems can be adjusted based on scanned cut edges and repeated optimization loops. Trumpf promises time and material savings as well as improved cut edge quality. 

More than 3% higher productivity – how digitalization in mechanical engineering increases productivity

The innovation momentum at HAWE and Trumpf is beyond question. But many smaller companies are also making strong progress with intelligent solutions without communicating this widely, emphasizes expert Schöllenhammer. “Such product-integrated solutions often disappear in broad digitalization debates behind the question of whether a company has already launched a major transformation program.”

The same applies to the use of AI. VDMA observes a much more concrete discussion than just a few years ago: Machine learning models can, for example, optimize production planning, anomaly detection, and technical documentation. “Many companies see the potential, but are still stuck in pilot phases,” says Guido Reimann, deputy managing director of VDMA Software and Digitalization. Here too, it is crucial not to treat AI as an isolated experiment, but to embed it in core processes and align it with measurable benefits. Schöllenhammer: “The bottleneck is no longer primarily in gaining insights. It’s about simply doing it.”

According to an IPA study, the use of digital technologies can deliver productivity gains of 0.2 to 3.3% per year. For smart factories, 2.8 to 4.4% per year is cited. Digital business models can increase value creation per employee by up to 20%. According to IPA, four fields together account for around 50% of productivity gains: modularization and software-defined production, the industrial metaverse, agentic AI and AI workflows, and the embodiment of AI.

Fraunhofer IPA is addressing the laggards: The biggest obstacles are a lack of organizational willingness to change, a lack of ability to integrate available technologies into existing processes, and missing technical foundations, especially data. Angela Graf from the Bavarian Research Institute for Digital Transformation (bidt) puts it in a nutshell: “Digital transformation is less a technological challenge than an organizational one.”